ANIMAL BEHAVIOUR, 2004, 68, 83e88doi:10.1016/j.anbehav.2003.07.017

Variation in mating system and group structure in two

populations of swift foxes, Vulpes velox

JAN F. KAMLER*, WARREN B. BALLARD*, PATRICK R. LEMONS* & KEVIN MOTE†

*Department of Range, Wildlife, and Fisheries Management, Texas Tech University, Lubbock

yTexas Parks and Wildlife Department, Pilot Point

(Received 21 May 2002; initial acceptance 6 August 2002;

final acceptance 21 July 2003; MS. number: A9365R2)

We studied 26 reproductive groups of swift foxes, Vulpes velox, from both high- and low-density areasduring three field seasons in northwestern Texas, U.S.A., to examine whether differences in populationdensity affect mating system and group structure. Although high- and low-density populations were onlyseparated by 40 km and vegetation and diets were similar between sites, polygynous groups, communaldenning and nonbreeding females occurred in the area of high density, whereas only monogamous pairsoccurred in the area of low density. Annual survival of adult swift foxes was 66% in the area of highdensity, but 44% in the area of low density. Predation from coyotes, Canis latrans, was the only mortalityfactor that differed (P ¼ 0:01) between sites and contributed most to differences in survival. Althoughprevious research indicated that variation in social systems among canids is related to bottomeup forces(i.e. food, habitat), the results of our study indicate that variation in social systems can also be related totopedown forces (i.e. predation, displacement by larger competitor).

� 2004 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

Polygamy is the predominant mating system of mostmammals, occurring in more than 97% of species studied(Kleiman 1977). The major exception occurs within thefamily Canidae, in which most species tend to be monog-amous (Kleiman 1977). Monogamy within Canidae prob-ably evolved in relation topair bonding andmale care of theyoung (Kleiman & Eisenberg 1973). However, there isconsiderable interspecific variation in mating systemsamong canids, and both polygamy and monogamy havebeen documented (Bekoff et al. 1981; Moehlman 1989;Geffen et al. 1996). Several factors are suggested tocontribute to the variation of mating systems among canidspecies, including body size (Moehlman1989) and resourceavailability (Geffen et al. 1996).Canids are also unique in that intraspecific variation in

mating systems may be as great as interspecific variation,whereby both polygyny and monogamy occur within thesame species (Moehlman 1989). Food availability, habitatavailability and resource dispersion have been suggestedas major factors contributing to intraspecific variationin reproductive strategy and group structure in canids

Correspondence and present address: W. B. Ballard, Department ofRange, Wildlife, and Fisheries Management, Texas Tech University,Lubbock, TX 79409, U.S.A. (email: warren.ballard@ttu.edu). J. F.Kamler is now at the Mammal Research Institute, Polish Academy ofSciences, 17-230 Bia1owie _zza, Poland.

830003e3472/03/$30.00/0 � 2004 The Association

(Macdonald 1983; Geffen et al. 1996). For example, thesocial system of golden jackals, Canis aureus, varies consid-erably with food dispersion and abundance (Macdonald1979a). Group sizes of grey wolves, C. lupus, and coyotes,C. latrans, often depend on prey size and availability(Bekoff & Wells 1980; Harrington et al. 1982; Messier &Barrette 1982). Both red foxes, Vulpes vulpes, and Arcticfoxes, Alopex lagopus, show intraspecific variation inmating system and group structure as a result of differencesin food and/or habitat resources (Macpherson 1969;Macdonald 1983; Hersteinsson 1984; Moehlman 1989).Mating systems have also been observed to change overtime within the same population, shifting from polygynyto monogamy, when food resources decline (red foxes:von Shantz 1984; Zabel & Taggart 1989).Among primate species, several factors have been sug-

gested to affect social organization, including food dis-tribution, risk of infanticide, conspecific competition,habitat dispersion, population density and predation risk(Wrangham 1980; van Schaik 1983, 1996; Dunbar 1988,1996; Moore 1999). Similarly, among avian species,several factors are suggested to affect mating systems,including male age, nesting dispersion, breeding synch-rony and breeding density (Birkhead & Møller 1992;Westneat & Sherman 1997). Under certain conditions,some of these factors might be more influential thanothers, but it is more likely that a combination of these

for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.

ANIMAL BEHAVIOUR, 68, 184

factors influences the social organization and matingsystem of a given population. Thus, although differencesin food abundance and habitat have been shown to affectintraspecific variation in reproductive strategies of canids,other factors such as mortality from predators might alsoaffect reproductive strategies of canids, at least undercertain conditions. In fact, some researchers have sug-gested that mortality from larger canids might contributeto variations in social systems of small canids (Voigt &Macdonald 1984; Cavallini 1996), although this hypoth-esis has not been tested.We studied two populations of swift foxes, separated by

40 km, in northwestern Texas, U.S.A., that differed indensity by at least two-fold. Initial results suggested thatdensity was related to differences in survival and mortalityfrom coyotes, and not to differences in food or habitatresources. This provided a unique opportunity to de-termine whether the mating system and group structureof swift foxes differed with respect to predation pressureand density.

METHODS

We studied 26 breeding groups of swift foxes (classified aswarranted, but precluded as endangered from 1995 to 2001under the Endangered Species Act; Clark 2001) at two sitesinnorthwesternTexas betweenAugust 1998 andMay2001.Site 1 was situated on Rita Blanca National Grasslands(RBNG) and adjacent private lands in west-central DallamCounty (36(31#N, 102(64#W). Site 2, approximately40 km east of site 1, was situated on a private cattle ranchon the border of Dallam and Sherman counties (36(24#N,102(19#W). Vegetation at both sites consisted of short-grass prairie, dominated by blue grama (Bouteloua gracilis)and buffalograss (Buchloe dactyloides), that was moderatelyto intensively grazed by cattle (Bos taurus). Although site 2was more fragmented with agricultural lands, swift foxesused the short-grass prairie more than 97% of the time(Kamler 2002). Coyotes were the primary predator of swiftfoxes at both sites, but the density of coyotes at each sitediffered (Kamler et al. 2003a). At site 1, the RBNG is openyear-round for hunting and trapping, but survivorship ofcoyoteswas high (90%: Kamler et al. 2003a). At site 2, ranchowners encouraged hunting of coyotes to reduce livestocklosses, and consequently, coyotes had low survival(54e56%: Kamler et al. 2003a).We captured swift foxes using box traps (Kamler et al.

2002). We initially concentrated our trapping effort nearthe centre of both study sites, then expanded outward ascaptureof unmarked foxesdecreased.Our researchprotocol(No. 00979BX) was approved by the Texas Tech UniversityInstitutional Animal Care and Use Committee. Swift foxeswere eartagged, radiocollared, and agedby toothwear, bodysize and reproductive condition (Rongstad et al. 1989). Weclassified foxes that were less than 10 months old asjuveniles (i.e. from March to mid-January each year) andclassified all other foxes as adults. We estimated the springdensity of swift foxes by calculating the minimum numberof adults that remained on each study site during May,whichwas themiddle of the pup-rearing period. The area of

each study site was determined by the total area encom-passed by all monitored foxes.

Foxes were considered to belong to the same familygroup if they used the same area and dens concurrently(Kitchen et al. 1999). Female foxes were considered to bebreeders if they were pregnant when captured, or showedevidence of nursing during or after the pup-rearing period.We defined a breeding group as polygynous if twobreeding females stayed in the same den (e.g. communaldenning) with pups during the entire pup-rearing period.One adult male was always associated with a pair ofbreeding females, although the male sometimes stayed ina separate den nearby. We defined breeding groupsconsisting of one breeding female and one adult male asmonogamous. We classified females as nonbreeders if theyused the same territory as that of breeding adults duringthe pup-rearing period, but showed no evidence of currentor former nursing. We compared mean group sizesbetween study sites using t tests (Zar 1996), and deemedresults significant when P!0:05.

We monitored den use, especially during the breeding(JanuaryeFebruary) and pup-rearing (MarcheJune) peri-ods, by radiotracking swift foxes at night and by visitingtheir underground burrows during the day. Throughoutthe study, we also recorded independent telemetrylocations (White & Garrott 1990) for study animals onceor twice each week, separated by at least 12 h. Weradiotracked from vehicles using null-peak systems, whichconsisted of dual, four-element Yagi antennas. For in-dependent locations, we radiotracked primarily during1800e0900 hours, when swift foxes were likely to be mostactive (Kitchen et al. 1999). We calculated locationestimates using the maximum likelihood estimationoption in the program Locate II (Pacer, Inc., Truro, NovaScotia, Canada). The mean error for reference collars(known locations) was 84 m (95% of errors were!145 m).

We determined annual home range sizes for adult swiftfoxes using the minimum convex polygon method, ascalculated by the Animal Movement program (Hooge &Eichenlaub 1997). We calculated home ranges for foxesusing a minimum of 30 locations and 6 months ofradiotracking data. Additionally, we pooled the locationsfor family members to calculate a total home range foreach family group. Group home ranges were calculatedonly if more than one individual per group met the abovecriteria. For individual group members, we calculatedhome range overlap based on the percentage of homerange area overlapping that of another family member. Wecompared mean home range sizes for individuals andfamily groups, and mean home range overlap, betweensites using t tests (Zar 1996), and deemed resultssignificant when P!0:05.

We estimated annual survival and cause-specific mor-tality rates for adult swift foxes using the programMicromort (Heisey & Fuller 1985). Causes of mortalityfor swift foxes were determined by necropsy. We classifiedswift fox deaths as coyote predation if fox carcasses hadhaemorrhaging and puncture wounds consistent withthat from coyote bite marks. We initially analysed the databy biological season to meet the assumption of constantsurvival (Heisey & Fuller 1985). Because preliminary

KAMLER ET AL.: VARIATION IN SWIFT FOX MATING SYSTEM 85

analyses indicated that survival did not differ betweenseasons or years, we grouped the data and comparedsurvivorship between sites using Z tests (Heisey & Fuller1985; Nelson & Mech 1986). Differences in survival andcause-specific mortality rates were deemed significantwhen P!0:05.

RESULTS

During the 3-year study, we radiocollared and monitored31 adult swift foxes on site 1 and 21 adults on site 2. Ofthese, we determined annual home ranges for 23 adults onsite 1 and 17 adults on site 2. Annual home range sizes(XGSE) for individuals were larger (Student’s t test:t38 ¼ 2:3, N ¼ 40, P ¼ 0:03) on site 2 (10.7 G 0.9 km2)than on site 1 (8.4 G 0.5 km2). Group home range sizesalso were larger (t7 ¼ �2:5, N ¼ 9, P ¼ 0:04) on site 2(16.6 G 2.4 km2) than on site 1 (11.3 G 0.3 km2). Homerange overlap of family members did not differ (t36 ¼ 0:7,N ¼ 38, P ¼ 0:46) between site 2 (73.1 G 4.0%) and site 1(78.8 G 6.5%).Spring densities of adult swift foxes were more than

twice as high on site 2 as on site 1 during all 3 years of thestudy (Table 1). We monitored 16 adult groups on site 1,and 10 adult groups on site 2. On site 1, all 16 adult groupswere monogamous pairs and no nonbreeding femaleswere present (Table 1). On site 2, three of 10 adult groupsconsisted of two adult females that denned communallythroughout the pup-rearing period, along with an associ-ated adult male that usually denned less than 30 m away.There were also four nonbreeding females (three yearlingsand one 2-year-old) present among the 10 groups. Overallmean group size was larger (t24 ¼ 3:4, N ¼ 26, P!0:01) onsite 2 than on site 1 (Table 1).Of the three pairs of adult females that denned com-

munally, we recaptured both females of one pair in earlyMarch when both were presumably in late pregnancy(based on unusually enlarged abdomens). Both of thesefemales were more than 2 years old and were similar in agebased on tooth wear. When captured prior to March, bothhad elongated and darkened nipples, suggesting that theyhad nursed the previous year (Mech et al. 1993). An

alternative explanation is that one of these females waspseudopregnant and did not actually give birth, aspseudopregnant canids also may have enlarged abdo-mens and can sometimes nurse (Asa & Valdespino 1998).However, pseudopregnant females in family groups aretypically yearlings that are reproductively suppressed byaggression from dominant females (Asa 1997). Becausedominant female foxes are highly aggressive towards sub-ordinates during breeding and birthing periods, breedersdo not allow subordinate females in the natal den duringthe first month after birth (Macdonald 1979b; von Shantz1984). Thus, we classified this group as polygynous be-cause both females showed evidence of pregnancy in earlyMarch and evidence of nursing the previous year, andboth females were several years old and similar in age, haddenned communally with pups throughout the entirebirthing and pup-rearing period (MarcheMay), and hadbeen associated with a single adult male. Females in thesecond pair were also more than 2 years old and similar inage based on tooth wear, and both females showed signsof nursing when they were captured the previous year.Both females also denned communally with pupsthroughout the entire birthing pup-rearing period andwere associated with an adult male. Because circumstanceswere similar to the first case, we assumed that bothfemales of this pair also gave birth and were a polygynousgroup. Females in the third pair were yearlings based ontooth wear and body size, and neither showed signs ofnursing the previous year when initially captured.Although neither female was recaptured during thepregnancy period, and thus, pregnancy could not beconfirmed, both denned communally with pups duringthe entire pup-rearing period and were associated with anadult male. Both of these females showed signs of previousnursing when captured the following autumn based onelongated nipples (nipples were not elongated in yearlingsthat did not nurse). Although one or both of theseyearlings could have been pseudopregnant and nursedanother’s pups without giving birth, we also assumed thatboth had given birth and were a polygynous group.Overall, the females within each polygynous group ap-peared to be of the same age, indicating that they mayhave been sisters.

Table 1. Adult spring density and group structure of swift foxes monitored at two study sites in northwestern Texasfrom 1999 to 2001

Spring density

(no. adult foxes/km2)

Monogamous

matings (N )

Polygynous

matings (N )

Nonbreeding

females* (N )

Mean group

size

Site 11999 0.09 5 0 0 2.02000 0.09 5 0 0 2.02001 0.11 6 0 0 2.0Summary 0.10 16 0 0 2.0

Site 21999 0.31 1 2 3 3.72000 0.19 3 0 1 2.32001 0.25 3 1 0 2.3Summary 0.25 7 3 4 2.7

*Nonbreeding females associated with reproductive groups.

ANIMAL BEHAVIOUR, 68, 186

During the study there were 16 confirmed adult mortal-ities on site 1, and sixon site 2.Overall, swift foxdeathswerecaused by coyote predation (N ¼ 14), vehicles (N ¼ 4) andother unknown causes (N ¼ 4). Annual survival was 66%on site 2, and 44% on site 1, although this difference wasnot statistically significant (Z test: Z ¼ 1:53, P ¼ 0:08).Cause-specific mortality rates differed statistically betweensites for coyote predation (Z ¼ 2:55, P ¼ 0:01), but not forvehicles (Z ¼ 1:47, P ¼ 0:10) or other unknown causes(Z ¼ 0:51, P ¼ 0:53).

DISCUSSION

Spring density of swift foxes was more than twice as highon site 2 as on site 1 during all 3 years of the study. Survivalwas also 50% higher on site 2 than site 1, and thus,differences in density were probably related to differencesin adult survival. Predation from coyotes was the largestcause of mortality overall, and the only mortality factorthat differed between sites, suggesting that coyote pre-dation strongly influenced differences in swift fox survivaland density between sites. There were fewer deaths fromcoyote predation at site 2 probably because the density ofcoyotes was lower as a result of increased rates of huntingand trapping of coyotes at this site (Kamler 2002). Inaddition to predation, coyotes spatially displaced swiftfoxes from their home ranges (Kamler et al. 2003b).Predation and spatial displacement are common amongcanid species, and can result in population suppression ofsmaller canids by larger canids (Johnson & Sargeant 1977;Peterson 1995; Crabtree & Sheldon 1999). Of all swift foxeskilled by coyotes during this study, none were consumed,similar to that reported by previous studies (Sovada et al.1998; Kitchen et al. 1999). This suggests that coyotes killedswift foxes for reasons other than food, such as competi-tion for food or territories.Differences in mating system and group structure of

swift foxes also occurred between study sites. In the area oflow density (site 1), all adult groups consisted of monog-amous pairs with no nonbreeding females. In contrast, inthe area of high density (site 2), 30% of all adult groupsconsisted of polygynous groups (two females and onemale), and nonbreeding females were present in 40% ofgroups. Because mortality still occurred on site 2, matingsystems and group structures there were not homogenousamong family groups. Our results are similar to those ofCovell (1992), who found that polygynous mating groupsamong swift foxes in Colorado occurred only when coyotenumbers were artificially reduced. During our study,coyote numbers were reduced on site 1 during 2000(Kamler et al. 2003a); however, this reduction occurredtoo late (January) to affect the breeding density of adultfoxes for that year. Among canids, communal denning oftwo breeding females was also reported for the closelyrelated kit foxes, Vulpes macrotis (Egoscue 1962), and redfoxes (Macdonald 1979b; Zabel & Taggart 1989), as well asbat-eared foxes, Otocyon megalotis (Nel et al. 1984), coyotes(Camenzind 1978) and grey wolves (Mech 1970).Several reasons might explain why the lower density of

swift foxes, due to predation by coyotes, decreased group

size and the occurrence of polygyny. First, high mortalitycreated vacant territories for both adult females andjuveniles to establish their own territories. Second, highmortality reduced the number of available females forboth communal denning and nonbreeding status. Voigt &Macdonald (1984) indicated that these same factors mighthave contributed to differences in mating system andgroup formation between two populations of red foxes.

The resource dispersion hypothesis predicts that groupformation of canids is dependent on heterogeneity offood and habitat distribution (Macdonald 1983). Similarly,differences in food resources are also thought to affectmating systems within canid species (Camenzind 1978;Hersteinsson 1984; Zabel & Taggart 1989). However,differences in food resources are unlikely to account forthe differences in mating systems and group structures onour study sites. Food items consumed by swift foxes onboth study sites included 53e68% insects, 38e49% smallrodents, 14e15% large rodents, 4e11% lagomorphs and7e15% birds (Lemons 2001). Because swift foxes areomnivorous and opportunistic feeders (Scott-Brown et al.1987; Sovada et al. 2001), their diets probably reflect thegeneral abundance of available food items (Sovada et al.2001). These relatively small differences in diet indicatethat food resources were relatively similar between sites.Thus, it is unlikely that differences in food resources con-tributed to the two-fold difference in density between sitesor major differences in mating system or group structure.

Differences in habitat resources are also unlikely toaccount for the differences in mating system and groupstructure between sites. Although we did not measurehabitat quality, short-grass prairie habitat, which histori-cally occurred in this region (Barbour & Billings 1988),dominated both study sites and was used by swift foxesmore than 97% of the time on both sites (Kamler 2002).Although more agricultural land occurred on site 2, homeranges of swift foxes were noncontinuous on both sitesdue to habitat availability and spatial displacement bycoyotes (Kamler 2002; Kamler et al. 2003b), indicatingthat foxes had limited access to available resources onboth sites. Additionally, den sites of swift foxes have beenfound in a variety of natural and human-altered habitats(Cutter 1958; Kilgore 1969), suggesting that den sites werenot a limiting factor on either site. However, limitingfactors related to resource dispersion could have occurredat a finer scale than that investigated during this study.Regardless, we believe that differences in mating systemand group structure of swift foxes in our study wererelated to the two-fold difference in density, which wasaffected most by differences in survival and coyotepredation. Thus, our study strongly suggests that ‘topedown’ factors, such as predation and spatial displacement,also can influence mating system and group structureof small canids. Due to the widespread suppression ofsmaller canids by larger canids (Johnson & Sargeant 1977;Peterson 1995; Cavallini 1996), influences of highmortality and spatial displacement on social systems ofsmall canids might be greater than previously believed.

The plastic social system of swift foxes might be anadaptation to the evolutionary constraints imposed ontheir spacing patterns and densities by larger canids

KAMLER ET AL.: VARIATION IN SWIFT FOX MATING SYSTEM 87

(Cavallini 1996). Thus, in areas with fewer coyotes, swiftfoxes may benefit by increasing mating opportunities viapolygyny, and by increasing group size to include non-breeding females, which might help to increase pupsurvival (Moehlman 1979). However, because density ofswift foxes was affected by mortality from coyotes, densityalone might have affected the mating system and groupstructure of swift foxes at each site. For example, amongavian species that are predominantly socially monoga-mous, breeding density is related to the use of alternativereproductive strategies within the same species (Westneat&Sherman1997; Richardson&Burke2001).Consequently,higher breeding densities in socially monogamousbird species increase the occurrence of polygamy, as thenumber of available mates increases (Westneat & Sherman1997; Richardson & Burke 2001). Similarly, populationdensity alone strongly influences social systems of primates(Moore 1999). However, other population parametersrelated to density, such as socioeconomic behaviours, alsoinfluence mating systems both in birds (Westneat &Sherman 1997) and in primates (Moore 1999). Thus,differences in the mating systems of swift foxes and othercanids probably result from a complex mix of interact-ing factors, both intrinsic and extrinsic, which expressthemselves according to each setof unique circumstances ata given time. The importance of various factors in theevolution of mating systems might be impossible todetermine, especially due to lack of knowledge about thespecific circumstances under which various social traitsevolved.

Acknowledgments

This project was supported by Texas Tech University, TexasParks and Wildlife Department, U.S. Forest Service,U.S.D.A.-Wildlife Services program, Kansas Departmentof Wildlife and Parks, BWXT Pantex, Houston ZoologicalSociety and Section 6 Grant E-1-12 from the U.S. Fish andWildlife Service’s Endangered Species Program. We thankC. C. Perchellet for assistance with the project. We alsothank the landowners in Sherman and Dallam counties,especially F. Pronger, for access to their property. This isTexas Tech University, College of Agricultural Sciences andNatural Resources technical publication T-9-942.

References

Asa, C. S. 1997. Hormonal and experimental factors in the

expression of social and parental behavior in canids. In: Co-

operative Breeding in Mammals (Ed. by N. G. Solomon & J. A.French), pp. 129e149. Cambridge: Cambridge University Press.

Asa, C. S. & Valdespino, C. 1998. Canid reproductive biology: anintegration of proximate mechanisms and ultimate causes.

American Zoologist, 38, 251e259.

Barbour, M. G. & Billings, W. D. 1988. North American Terrestrial

Vegetation. Cambridge: Cambridge University Press.

Bekoff, M. & Wells, M. C. 1980. The social ecology of coyotes.

Scientific American, 242, 130e148.

Bekoff, M., Diamond, J. & Mitton, J. B. 1981. Life-history patterns

and sociality in canids: body size, reproduction, and behavior.

Oecologia, 50, 386e390.

Birkhead, T. R. & Møller, A. P. 1992. Sperm Competition in Birds:

Evolutionary Causes and Consequences. London: Academic Press.

Camenzind, F. J. 1978. Behavioral ecology of coyotes (Canis latrans)on the National Elk Refuge, Jackson, Wyoming. Ph.D. thesis,

University of Wyoming.

Cavallini, P. 1996. Variation in the social system of the red fox.

Ethology, Ecology, and Evolution, 8, 323e342.

Clark, J. R. 2001. Endangered and threatened wildlife and plants;

annual notice of findings on recycled petitions. Federal Register,66, 1295e1300.

Covell, D. F. 1992. Ecology of the swift fox (Vulpes velox) insoutheastern Colorado. M.S. thesis, University of Wisconsin,

Madison.

Crabtree, R. L. & Sheldon, J. W. 1999. Coyotes and canid

coexistence in Yellowstone. In: Carnivores in Ecosystems: the

Yellowstone Experience (Ed. by T. M. Clark, A. P. Curlee, S. C.

Minta & P. M. Kareiva), pp. 127e163. New Haven, Connecticut:Yale University Press.

Cutter, W. L. 1958. Denning ecology of the swift fox in northernTexas. Journal of Mammalogy, 39, 70e74.

Dunbar, R. I. M. 1988. Primate Social Systems. Ithaca, New York:Cornell University Press.

Dunbar, R. I. M. 1996. Determinants of group size in primates:a general model. In: Evolution of Social Behaviour Patterns in

Primates and Man (Ed. by W. G. Runciman, J. Maynard Smith &

R. I. M. Dunbar), pp. 33e57. Oxford: Oxford University Press.

Egoscue, H. J. 1962. Ecology and life history of the kit fox in Tooele

County, Utah. Ecology, 43, 481e497.

Geffen, E., Gompper, M. E., Gittleman, J. L., Luh, H., Macdonald,D. W. & Wayne, R. 1996. Size, life-history traits, and social

organization in the Canidae: a reevaluation. American Naturalist,147, 140e160.

Harrington, F. H., Paquet, P. C., Ryon, J. & Fentress, J. C. 1982.Monogamy in wolves: a review of the evidence. In: Wolves of theWorld (Ed. by F. H. Harrington & P. C. Paquet), pp. 209e222. Park

Ridge, New Jersey: Noyes.

Heisey, D. M. & Fuller, T. K. 1985. Evaluation of survival and

cause-specific mortality rates using telemetry data. Journal of

Wildlife Management, 49, 668e674.

Hersteinsson, P. 1984. The behavioral ecology of the Arctic fox

(Alopex lagopus) in Iceland. Ph.D. thesis, Oxford University.

Hooge, P. N. & Eichenlaub, B. 1997. Animal Movement Extension

to Arcview. Version 1.1. Anchorage: Alaska Biological Science

Center, U.S. Geological Survey.

Johnson, D. H. & Sargeant, A. B. 1977. Impact of red fox predation

on the sex ratio of prairie mallards. United States Fish and WildlifeService Research Report, 6, 1e30.

Kamler, J. F. 2002. Relationships of swift foxes and coyotes innorthwest Texas. Ph.D. thesis, Texas Tech University.

Kamler, J. F., Ballard, W. B., Gilliland, R. L. & Mote, K. 2002.Improved trapping methods for swift foxes and sympatric coyotes.

Wildlife Society Bulletin, 30, 1262e1266.

Kamler, J. F., Ballard, W. B., Gilliland, R. L., Lemons, P. R. & Mote,K. 2003a. Impacts of coyotes on swift foxes in northwestern Texas.

Journal of Wildlife Management, 67, 317e323.

Kamler, J. F., Ballard, W. B., Gilliland, R. L. & Mote, K. 2003b.Spatial relationships between swift foxes and coyotes in north-

western Texas. Canadian Journal of Zoology, 81, 168e172.

Kilgore, D. L. 1969. An ecological study of the swift fox (Vulpes

velox) in the Oklahoma Panhandle. American Midland Naturalist,81, 512e533.

Kitchen, A. M., Gese, E. M. & Schauster, E. R. 1999. Resourcepartitioning between coyotes and swift foxes: space, time, and

diet. Canadian Journal of Zoology, 77, 1645e1656.

ANIMAL BEHAVIOUR, 68, 188

Kleiman, D. G. 1977. Monogamy in mammals. Quarterly Review of

Biology, 52, 39e69.

Kleiman, D. G. & Eisenberg, J. F. 1973. Comparisons of canid and

felid social systems from an evolutionary perspective. Animal

Behaviour, 21, 637e659.

Lemons, P. R. 2001. Swift fox and coyote interactions in the

short-grass prairie of northwest Texas: competition in diets andden site activity. M.S. thesis, Texas Tech University.

Macdonald, D. W. 1979a. The flexible social system of the goldenjackal, Canis aureus. Behavioral Ecology and Sociobiology, 5, 17e38.

Macdonald, D. W. 1979b. Helpers in fox society. Nature, 282,69e71.

Macdonald, D. W. 1983. The ecology of carnivore social behaviour.Nature, 301, 379e384.

Macpherson, A. H. 1969. The dynamics of Canadian Arctic foxpopulations. Canadian Wildlife Services Report Series, 8, 1e49.

Mech, L. D. 1970. The Wolf: Ecology and Social Behavior of an

Endangered Species. New York: Natural History Press.

Mech, L. D., Meier, T. J. & Seal, U. S. 1993. Wolf nipple measure-

ments as indices of age and breeding status. American MidlandNaturalist, 129, 266e271.

Messier, F. & Barrette, C. 1982. The social system of the coyote(Canis latrans) in a forested habitat. Canadian Journal of Zoology,

60, 1743e1753.

Moehlman, P. D. 1979. Jackal helpers and pup survival. Nature,

277, 382e383.

Moehlman, P. D. 1989. Intraspecific variation in canid social

systems. In: Carnivore Behavior, Ecology and Evolution. Vol. 1 (Ed.

by J. L. Gittleman), pp. 143e163. Ithaca, New York: Cornell

University Press.

Moore, J. 1999. Population density, social pathology, and behav-

ioral ecology. Primates, 40, 5e26.

Nel, J. A. J., Mills, M. G. L. & van Aarde, R. J. 1984. Fluctuatinggroup size in bat-eared foxes (Otocyon megalotis), in thesouthwestern Kalahari. Journal of Zoology, 203, 294e298.

Nelson, M. E. & Mech, L. D. 1986. Mortality of white-tailed deer innortheastern Minnesota. Journal of Wildlife Management, 50,691e698.

Peterson, R. O. 1995. Wolves as interspecific competitors in canid

ecology. In: Ecology and Conservation of Wolves in a Changing World

(Ed. by L. N. Carbyn, S. H. Fritts & D. R. Seip), pp. 315e323.

Edmonton, Alberta: Canadian Circumpolar Institute.

Richardson, D. S. & Burke, T. 2001. Extrapair paternity and

variance in reproductive success related to breeding density in

Bullock’s orioles. Animal Behaviour, 62, 519e525.

Rongstad, O. J., Laurion, T. R. & Anderson, D. E. 1989. Ecology of

Swift Fox on the Pinon Canyon Maneuver Site, Colorado. FortCarson, Colorado: Final report, Directorate of Engineering and

Housing.

van Schaik, C. P. 1983. Why are diurnal primates living in groups?

Behaviour, 87, 120e144.

van Schaik, C. P. 1996. Social evolution in primates: the role of

ecological factors and male behaviour. In: Evolution of Social

Behaviour Patterns in Primates and Man (Ed. by W. G. Runciman,

J. Maynard Smith & R. I. M. Dunbar), pp. 9e31. Oxford:Oxford University Press.

Scott-Brown, J. M., Herrero, S. & Reynolds, J. 1987. Swift fox. In:Wild Furbearer Management and Conservation in North America

(Ed. by M. Novak, J. A. Baker, M. E. Obbard & B. Malloch),

pp. 433e441. Ontario, Canada: Ministry of Natural Resources.

von Shantz, T. 1984. ‘Non-breeders’ in the red fox Vulpes vulpes:

a case of resource surplus. Oikos, 42, 59e65.

Sovada, M. A., Roy, C. C., Bright, J. B. & Gillis, J. R. 1998. Causesand rates of mortality of swift foxes in western Kansas. Journal of

Wildlife Management, 62, 1300e1306.

Sovada, M. A., Roy, C. C. & Telesco, D. J. 2001. Seasonal foodhabits of swift foxes (Vulpes velox) in cropland and rangelandlandscapes in western Kansas. American Midland Naturalist, 145,101e111.

Voigt, D. R. & Macdonald, D. W. 1984. Variation in the spatial and

social behaviour of the red fox, Vulpes vulpes. Acta Zoologica

Fennica, 171, 261e265.

Westneat, D. F. & Sherman, P. W. 1997. Density and extra-pair

fertilizations in birds: a comparative analysis. Behavioral Ecology

and Sociobiology, 41, 205e215.

White, G. C. & Garrott, R. A. 1990. Analysis of Radio-tracking Data.

San Diego: Academic Press.

Wrangham, R. W. 1980. An ecological model of female-bonded

primate groups. Behaviour, 75, 262e300.

Zabel, C. J. & Taggart, S. J. 1989. Shift in red fox, Vulpes vulpes,

mating system associated with El Nino in the Bering Sea. AnimalBehaviour, 38, 830e838.

Zar, J. H. 1996. Biostatistical Analysis. 3rd edn. Upper Saddle River,

New Jersey: Prentice Hall.